Great insightful video as I recently purchased an fx6

I have 5 FX3s with many productions going to broadcast. I leave all the cams on XAVC HS for everything I shoot and just call it a day. Files sizes are amazingly small and the quality is superb. I have not shot a single frame of video in INTRA mode and never will with these cams. No need.

Thanks for this Video it has confirmed what I had long suspected to be the case.

@deadendboredom

Жыл бұрын

Make sure you also watch the second part for more examples and tests In different situations.

Good video. I use H265 with Proxies on my A7SIII and All-Intraframe codecs with my FX6. In most cases I operate with clients that generally prefer All-Intraframe codecs for fast turnarounds. I'd also mention the complaints you have are also very difficult to discover with clientele and in my opinion the only people that would ever notice that 1% problem would be pixel peepers (you and I). The issue is most likely going to be even more moot once it reaches onto social media where the images gets even more muddied or destroyed in the compression process. Do I like running an Atomos recorder? Sure. Do I want to use my Atomos for run and gun work? Hell no. I think the All-Intraframe codecs are plenty for work. If a client wants an even higher data rate (Which most don't have the money for) they'd be asking us to run Arri Alexa's with ProRes 4444 over having to run the Atomos recorder with ProRes or even PRR.

@deadendboredom

Жыл бұрын

Thanks for your comment, I’m all with you with everything you write. But I would like to add one more thing, because for me it goes beyond pure pixel peeping. For me, someone peeps pixels when he looks and nags around everywhere for no reason and for fun, even though it has no practical relevance. Globally, it may actually only affect a very small percentage, but I would like to distance myself a little from that. For example, I work a lot with green screen and the difference is blatant - it is unclear to me how Sony can call a camera a cinema camera and not meet important prerequisites for it. If I make a Hue vs Hue color change in grading, large colored pixel blocks appear in all-i 422 10bit sometimes faster than in 8bit 420 Long GOP. If sony didn’t classify the camera in a cinema line, I would be much more gracious with the criticism and would simply accept that the quality is intended for broadcast and social media. It’s such a pity that Sony actually has the great solution here for those who also want to enjoy autofocus on the big screen. And for all those who now say that a cinema camera does not need autofocus: you are right, the Internet will never prevail and digital will never replace analog. If an autofocus works reliably, then it can in no case hurt to have this additional option. I can well imagine that it would be much more widespread if the mindset „thats how it is, it has always been the case, it must stay that way“ did not exist. For settings where you want to focus manually for artistic reasons, you can still do that.

Great video thanks!

This is really fascinating to see. I’ve tried All-I recordings on several different mirrorless cameras from multiple manufacturers now, and I’ve pretty much always been underwhelmed by the image. But the general narrative has pretty much always been that it’s better than LongGOP, I’ve just never had the time or means to thoroughly test those claims. It always made since that for shots with a lot of movement like trees blowing in the wind, that it would probably benefit from All-I but again, I’ve never really been able to confirm that in practice and I’ve never been impressed with ALl-I codecs on mirrorless cameras. At least so much to choose it over Long GOP. And even in some cases I’ve personally felt the image wasn’t as strong with ALL-I. Your video seems to almost tangentially confirm my suspicions about the codec; at least enough that I’ll always be a bit suspicious of using it without hard confirmation that it is in fact a better image on a camera to camera basis. In the meantime, I’ll keep leaning on ProRes if I’m not trying to save space with Long GOP. Thanks for doing these tests!

@deadendboredom

Жыл бұрын

In the second part, I delve into more examples, and in the video description, you'll find a download link to the clips if you want to compare them yourself.

정말로 유용한 정보를 나눠주었다! 감사!

@deadendboredom

Жыл бұрын

당신은 매우 환영하고 댓글을 주셔서 감사합니다!

ProRes/DNx and Raw are industry standards for a reason. You need a higher bitrate all-i codec for serious post production. Something like 400 mbits minimum but preferably 600-800 mbits. It’s cool to see some new cameras from Panasonic and Fuji start to include ProRes in the cameras and hopefully Sony will do this to.

Someone test the canon 605 4k settings , I updated my camera then went strait to Europe and noticed a very noise image in highest image setting. Not around my camera to look right now

Interesting, I ve had the same problem with XAVC ALL-I, specially in blacks

Try shooting shots with random motion such as rippling water, leaves blowing in the wind, crops moving or even crowds of people moving about and you'll see the quality of long gop drop like a brick. Netflix and pretty much every broadcaster in the world is perfectly happy with XAVC-I because it is consistent no matter what you shoot and delivers high enough quality for most applications that cameras at this price point are used for. They don't accept XAVC-S or L because there are so many scenes that cause it serious issues.

@alisterchapman1556

Жыл бұрын

Pans do not stress long gop codecs particularly hard as everything moves in a Predictable way, so the prediction P frames deal with it. One tree with randomly fluttering leaves in a shot can eat up most of the encoding bit rate in a long gop codec causing the quality of the entire frame to drop.

@deadendboredom

Жыл бұрын

@@alisterchapman1556 I have heard this statement several times now and I don’t quite understand where the impression arises that I demand that people should exclusively use Long GOP. My wish simply is that we get an even better All I Codec. Didn’t I make it clear enough in the video that everyone should do their own tests to find the best recording modes for their work? I am serious about this question, if I have conveyed a false picture with special statements, then I take the criticism to heart. So tell me where and why you get that Impression.

@alisterchapman

Жыл бұрын

​@@deadendboredom Umm.... "Sony's ALL I Quality Problem" or "XAVC-I is best quality is simply not true" You have only looked at one single very particular aspect of image quality - mosquito noise. And you will almost always have more mosquito noise in an I frame codec than a long gop codec because the long gop file structure cannot deal with things that are different in every frame. You have not looked at resolution loss, posterisation, vector noise, motion artefacts (except a ridiculous and un representative pan), you haven't looked at how noise affects long gop codecs. There isn't a quality problem with XAVC-I. It's a compressed codec and it isn't perfect, but it isn't broken either, there isn't a "problem" with it. Do you honestly think that organisations like the Netflix or the BBC (who's highly skilled engineers do an immense amount of testing) would have approved XAVC-I but not XAVC-L it if there was a quality problem with XAVC-I? In the world of film and TV production a codec that delivers variable image quality depending on what you are shooting is not generally deemed acceptable because of the inconsistencies this introduces, you never know how your footage will look. You can't send a crew out to cover stories but then tell them they can't shoot water or trees on a windy day because it will look bad. And you can't randomly keep switching codecs during a production, because you aren't going to see these teeny, tiny differences in the field. Additionally, long gop codecs include motion artefacts, I frame codecs do not. You have only looked at a single type of artefact, which is mosquito noise. You have not looked at vector and motion artefacts which in a video production chain from camera to screen get multiplied through concatenation. The fact that you consider the 8 bit XAVC-L from the FX6 to be better than the 10 bit XAVC-I shows how limited your testing actually is or that you're quality criterion is dependent on just one single factor rather that the overall bigger picture.

@deadendboredom

Жыл бұрын

@@alisterchapman I don’t know how often I will respond to your questions and statements in detail. I didn’t advise anyone to record XAVC L, I even briefly showed the horrible artifacts in motion with Long GOP. I would like to simply see another All I option with a higher Bitrate And I personally would have liked to know before the purchase that the internal recording does not meet my requirements. Are you happy if I change the title and thumbnail? What would you say about how we get out of this situation positively?

@CLIFFLIX

Жыл бұрын

Alister- I dont understand Sony's codec choices. I cant wrap my head around why they chose the numbers they did and locked them down and to not allow more choices. 30 frame with one choice of 300mbps? Or...the FX6 with no 10bit long GOP options? Or...FX3/30 and other new Alphas with HEVC that gives you 24 and 60 frame....but blocks out 30p? Or...h.264 Long GOP bolted down to 140mbps in 30p. You know Sony better than anybody. Why do they make these bizare codec choices?

That is one of the main reasons why prefer Canon cameras. With their XF-AVC 410Mbit/s All-I you don't get the blocking compression artifacts AND the better editing performance.

@deadendboredom

Жыл бұрын

Good to know, thanks for your comment! A Canon camera from the C series? Could you possibly record a subject with many details with a very wide-angle lens and send it to me? Once with the ALL-I and once with Long GOP? That would help me a lot!

Is it fair to say that XAVC-I has roughly the image quality as ProRes LT? It seems to me to be very similar in final result.

@deadendboredom

Жыл бұрын

ProRes LT is a bit better than XAVC-I, the pixel blocks are less visable in comparison.

@CLIFFLIX

Жыл бұрын

@@deadendboredom My question for Sony is why 300mbps was chosen for 30p? Why?....what was the technical justification for it? Media/card limits? Hardware encoder limit? I understand having that low all-I bitrate "available" as a "choice"....but not the "only" choice. Why in the world would you not have other rates to choose from in 2022? I can tell you that Panasonic's 400 all-I is fairly bulletproof. You can bend it, stretch it and fold it into a pretzel and its tough as nails. I REALLY dont understand why Sony is so stubborn on this. The FX6 long GOP is locked to 8bit also! The FX3/Fx30 and all other Alphas even have 30p HEVC removed! They allow 24 and 60 frame but block out 30p....stupid

@solodagci

3 ай бұрын

@@CLIFFLIX So much this. Sony arbitrarily removes certain key options from their cameras. Never give an explanation and never add them later via firmware updates. My next camera upgrade certainly won't be Sony. I'm waiting for next big thing from Panasonic.

Thank you for the information. Can you please provide the (ALL-I and Long GOP) orginal video files?

@deadendboredom

Жыл бұрын

Yes, I can. You will now find a download link in the video description with the material from the FX3, FX6 and the ProRes recordings. I'm looking forward to your findings.

@automatisatorpeace5453

Жыл бұрын

@@deadendboredom Thank you :)

What you're concerned about falls into the classification of "if it really matters - get a bigger budget" Since Sony introduced XAVC-I, millions of hours of footage have been captured and used in professional productions from corporate to high end documentary and it will rarely have caused an issue. If it matters to you, use another camera and lose the Sony benefits, hire a focus puller, or use the external recorder and accept the extra steps and rigging required. Operators who need to work as fast as possible don't have time for that, just to have a small increase in image quality.

@luissisinno6815

Жыл бұрын

Word!

@deadendboredom

Жыл бұрын

I understand the point you make, but I'm not a friend of solving problems by throwing a lot of money on it - especially if they are artificial hurdles, as in this case. The memory cards are fast enough, at higher frame rates the bit rate is higher and good quality is provided via HDMI. I can understand in the short term why a camera manufacturer artificially restricts its "cheaper" cameras and they have the right to do what they want, but then I also have the right to educate and ask for change. Why does the desire for better ALL-I Bothers you, would it take something away from you?

@andrewtaylor5103

Жыл бұрын

@@deadendboredom I can't work out what you're even talking about in that comment. It isn't an artificial hurdle. Manufacturers have to balance the capability of the tech with cost, power consumption, usability, performance, probability of failures, all restricted by the laws of physics and within a specification that the majority of their customers need to do their jobs and are willing to pay for. Very few of Sony's customers will have any benefit by recording Prores or raw compared to internal XAVC-I. The examples you show are not indicative of real world production. I already burn through 200-300 gigs on an average day. I'm not interested in recording more data than I need to. That already takes long enough to transfer at the end of a shoot. Sony already have a "better" All-I codec, XAVC Class 480, but it's not available in the FS/FX series - which I'm sure we all agree is unfortunate. I'm curious why you didn't make that the leading point of your video? (I suspect the reason it's not included is a hardware tradeoff, and customers who need more than standard XAVC-I can go the route of an external recorder) In another comment it's apparent you didn't know that 4k XAVC-L (long gop) is only 8 bit. Do you really have the experience to make a video about this topic, possibly misleading newcomers down the wrong path? I'm guessing you don't, otherwise you would also have included in your video examples of where XAVC-L is a SIGNIFICANTLY worse in real world professional settings. Not just a few seconds of rapid panning. For anyone else reading, to give some more insight in response to this video - XAVC-L is usually fine for scenes with little movement like long interviews, speeches, scenes in which the camera is static or has slow movement, and where you don't need the extra color grading flexibility of XAVC-I and want to reduce file size. For pretty much anything else, XAVC-I is going to hold up better, both in overall image quality as well as grading flexibility. Surfing, football, any sport, concerts, fashion, fast detailed in-frame movement or camera movements will benefit from XAVC-I.

@deadendboredom

Жыл бұрын

@@andrewtaylor5103 I understand that it is beneficial to offer a balanced codec, but why not an additional option for customers who want more? You said yourself it is unfortunate. I'm all with you that it only affects a small percentage, but I'm convinced you should know what you're buying. If I had known this before buying, I would have considered again whether there were more suitable alternatives. Unfortunately, you can't know or research everything beforehand. The static images are not representative of all real-world scenarios is correct - what about the green screen and bluescreen recordings? Nobody does that, so it doesn't matter? 8bit XAVC L: my main focus was on the FX3 and I didn't claim that the FX6 had 10bit, I don't quite see how Thies is misleading. But you're right, I didn't explicitly check how many bits the codec has, as it wasn't so relevant for the test. As I already answered under another statement from you: I never demanded that all people should be shooting in Long GOP, quite the contrary. Again, I can only repeat what I said often enough in the video: everyone should do their own tests and see which codec is the best option for their work. If I haven't communicated this clearly enough, then I'm sorry. Would you be happy if I deleted the video and uploaded a revised version where it is clearly stated that my wish is that the Class 480 codec would be included in the FX series, because there are also users who expect more quality from a "Cinema line"? Also with the clear indication that the current situation affects the fewest actively and that the testshots exaggerate the Phenomenon?

@alisterchapman

Жыл бұрын

@@deadendboredom How is not knowing not knowing the bit depth of a codec not relevant in a video about codec quality? At 2.40 you show the FX6 menu and you show switching from XAVC-I to XAVC-L, you are implicitly advising viewers to use XAVC-L on the FX6, but you clearly either haven't done your homework or haven't actually tested the codec. Why talk abut something you haven't actually tried or tested? And if you have tested, what exactly is you quality criterion based on? Only mosquito noise? Do posterisation, resolution loss, motion vectors not count as image quality?

I find this concerning! What also annoys me is that you can't shoot 25p in 4k XAVC HS h.265 on the FX3's. So my only option as a PAL user is to shoot H.264 which is fine, but really annoying if you doing a very long interview with multiple cameras because it's 140MB, when 24p gets an equivalent 100MB H.265 option. I feel cheated out of 40MB a sec space saving!

@alisterchapman1556

Жыл бұрын

Lower frame rates and H265 suffer from greater encoding artefacts because the length of time between I frames increases, so the motion differences are greater which means breakup of the image more likely when shooting anything with motion. To get similar image quality at 25fps the bit rate would need to be similar to the 50fps.

Also worth noting (another thing i intensely dislike about Sony)... The FX3 XAVC SI codec saves space and is still 422 10-bit. However on the FX6 the XAVC-L is only 8-bit. Sony...please explain why I can't shoot a compressed 10-bit image on a more 'professional' camera such as the FX6???

@deadendboredom

Жыл бұрын

I understand your frustration and I also don’t understand why Sony doesn’t offer a 10bit long GOP in FX6, but I would like us to be careful with the word „hate.“ Interestingly, I have already recorded various test clips and compared 10Bit All-i with 8Bit Long GOP and noticed surprisingly little difference. Even if I make completely exaggerated and unrealistic color shifts and corrections, the image sometimes gets destroyed earlier in All I, thanks to the already known pixel blocks. So here, too, it’s a bit of a case of: the pure numbers and specifications make us believe that something is good or bad, but the reality is a little different. P.s. Funnily enough, 8Bit even has the advantage that it runs smoother on Editing, since most graphics cards can accelerate 8bit 420, but not 10bit 422.

@InvisageStudios

Жыл бұрын

@@deadendboredom hate was the wrong word. Incredibly frustrating maybe better. I’ve accidentally filmed in 8 bit on the FX6 as I had no idea the LongGOP wasn’t 10-bit. Kinda got my butt kicked by an agency for it 😂 but like you said - after applying the s709 LUT… no one could really tell, it was exposed correctly so no extreme correction was required and it held together just fine. Especially for social media delivery. I’ve since filmed ALL-I but have had to invest hundreds more $$ in media because of how much data it chews. I’ve stuck with V90 SDCards… but once again - Sonys frustrating feature blocks the FX3 from higher frame rates even though the exact same cards work flawlessly on the FX6 (I’ve found AngelBird v90 to be incredible). Sony in my opinion is doing everything else right. It’s just these little quirks that throw your workflow a bit when you’d expect the “cinema line” of cameras to all be somewhat similarly functionally and capability wise.

@deadendboredom

Жыл бұрын

@@InvisageStudios for Typ-A i can recommend the Lexar CFexpress. It works fine with the FX3 in 120p and is kind of affordable compared to V90 cards - at least in Germany I can currently buy 160GB for little over 200€.

Hello how are you, I know it has nothing to do with this topic, but seeing that you have an fx6, you could still help me.. I am trying to attach the proxies files that create my fx6 in premier pro but I get this error (proxy media and full resolution media should have the same audio channels), the thing is that the original file has 8 channels and the proxy 2, I have tried to change it in the camera, in the proxy menu and it won't let me change the audio channels, I have also tried to do it in the premier pro menu and the same thing keeps coming out. Does anyone know something?? can you help me

@deadendboredom

Жыл бұрын

Unfortunately, I don't have a solution for you. I also don't get the proxy clips in Resolve relinked. It's a bit obscure to me anyway why the normal clips have 8 channels at all, since you can't control or use channel 5-8 anywhere. I just looked through all the options again and couldn't find any setting options to reduce or adjust the number of audio channels. One thing I noticed: for me, the proxy recordings had a slightly different file name at the end - but it also didn't work after adjusting the names. This sounds like a concern that should definitely be addressed to Sony, because in its current form, the proxy recordings, if we have not overlooked something, make no sense.

Is it possible to use the black magic external recorder with the FX3/FX6?

@deadendboredom

Жыл бұрын

Absolutely - you just can’t record RAW, only ProRes.

i have only two questions, why are you on 1/640 shutter speed and Iso 1250 in Slog3?

@deadendboredom

Жыл бұрын

The short clip you’re talking about actually has some unexpected settings. At first, it has nothing to do with the test and comes from another video about autofocus. All settings of the tests are recorded (except for lowlight shots) with iso 800. But you are addressing an interesting point, because I first asked myself „what, where did I record with these funny Settings“. In fact, it actually happens to me comparatively often that I turn the wrong dial on the FX3 (aperture, iso, shutter) or touch the wheel completely by mistake... the FX6 is much more pleasant, as an accidental change is less likely there. With AF videos, in addition, another person who is not familiar with cameras presses start and stop and is even more likely to accidentally change settings. By the way, in the second video, which is about a more precise comparison of the codecs, there is an explicit difference between short shutter times and normal 180 degrees. P.s. good eye that you noticed the short scene, next time with such a question briefly link the time so that I can find it quickly. Cheers!

@Dhatboyaj

Жыл бұрын

@@deadendboredom Understood but honestly i dont think you did anything wrong there, i believe if theres no ND filter at the time of recording outside and you dont want the highlights blown out you can use that setting you have or lesser so yh , but great video tho PS; Got it i will check that out, and haha I am a cinematographer with fx3,and fx6 . So i learn everyday on the go.haha

@Dhatboyaj

Жыл бұрын

@@deadendboredom Also i dont see the ALL - I or XAVCS- I to be bad but i will run some tests as well

Isn’t long gop codec on the fx6 8bit when filiming in 4K though? And All I is 10bit.

@deadendboredom

4 ай бұрын

You're not the first person to notice this, and you're right. I was surprised myself, as I wouldn't have seen the difference without knowing it and looking for it specifically. Unlike the FX3, there was no indication of this in the FX6 menu, and without challenging scenes recorded plus an assessment on a 10-bit display, the difference between 8-bit and 10-bit is surprisingly small in my opinion.

Using FX6 in studio that has perfect lighting is it over built or too much? Or just FX3 is good enough? For grading

@deadendboredom

2 ай бұрын

The two cameras have the same sensor, and the image is nearly identical. If you don't need the ND filter (which I assume Is the Case in a studio with controlled lighting), then the FX3 is a good choice. The FX3 even offers more codec options.

@oupekha

2 ай бұрын

@@deadendboredom Thanks a lot

Have you tried recording 4k 120 fps in S&Q mode? It records at 1200 Mbps All intra. The image seems to be way better

@deadendboredom

2 ай бұрын

I have, but the bitrate per frame remains the same, so the quality won't be much better. 300 Mbps at 30p and 1200 Mbps at 120p.

4k Long GOP is 8bit on the FX6. I would rather have 10bit with All-I.

@deadendboredom

Жыл бұрын

I haven't checked that before, but you're right! In general, the FX6 has fewer codecs - h.265 doesn't seem to exist at all, if I haven't overlooked something.

@luissisinno6815

Жыл бұрын

@@deadendboredom LONG GOP is 8bit 4.2.0. XAVC I is 10bit 4.2.2. If you care about semantics, this means that XAVC is higher quality.

@deadendboredom

Жыл бұрын

@@luissisinno6815 You actually caught me - good that you probably never overlook anything you can't immediately see. Then let's briefly expand the semantics: on paper you are absolutely right, the numbers suggest superior quality, but quality does not take place on paper for me, but in visible quality. Would you say XAVC I is optically better?

@luissisinno6815

Жыл бұрын

First of all, you are building your whole argument on the "quality" of a still image, to draw conclusions about a camera whose function is to capture moving images. This makes no sense. If I really want "more quality" in a still image, I use a still camera, not a video camera.

@Tigerodoes

4 ай бұрын

It's 10bits of pure noise. Totally ungradeable footage... the FX6 is a dslr video camera with sdi out 😂

Dude, are you shooting 1 stop over? and pulling down in post? Because it's pretty much flawless for me if you shoot like that.

@deadendboredom

Жыл бұрын

No, I record with ISO 800 and measured with lighmeter and exposed Accordingly. The colors and exposure you see are complete without adjustments and over or under exposure. The deficits I am addressing here are mainly visible when you zoom very close or have advanced claims for special effects such as chroma keying. For example, if you only make videos for social media, all this will be less relevant to you. In the second Video, I go into more detail With more different test scenes. You can even download the original clips in the video description to check for yourself.

@antonstafeyev3606

2 ай бұрын

@@deadendboredom Ok so here is the problem, if you are shooting LOG, you overexpose a shot like a stop over i even do two stops. Second you need to understand what log is. it is a logarithmic function that maps data evenly because, well, linear storage would waste so much space and fitting 13 stops of DR into 10-bit codec would be impossible. So LOG function basically drags your shadow data up a bit, spreading it evenyl with highlights and stuff, So in order to get better shadow representation you need to overexpose a bit. this is what CIne EI is for. you expose for 400 ISO with your aperture and ND but shoot at 800 ISO. When you convert that into nice rec 709 you retain a lot more information in shadows but not as much in highlight. You would do the opposite to retain highlight info.

@deadendboredom

2 ай бұрын

@@antonstafeyev3606 I am aware of what LOG does and that dark areas in the image are susceptible to compression artifacts. The reason I made this video was the assertion by many KZreadrs that the best codec and quality for Sony is all-I. I believe I was able to show in both the first and this video that this is not the default case. Yes, overexposing helps to reduce not only the artifacts but also the noise, but this also often works better in Long GOP. Wouldn't you also wish that Sony offered a higher bitrate for all-I? Aren't you a bit disappointed by the codec of a "cine camera," especially if you are used to R3D and other higher-quality codecs?

@antonstafeyev3606

2 ай бұрын

@@deadendboredom No no, thats actually a defined workflow by sony, and introduction of CineEI is a way to enforce it. I actually bought Fx6 to shoot non raw stuff with autofocus. but was amazed at a quality i get out of it. Regarding your issue with XAVC i , i ran a test and underexposed while correcting exposure in post and that is not as bad as your footage. Yeah more blocking due to tighter packed information but not as bad as yours. Try shooting at 12800 ISO and setting EI to 3200. The way compression works is by merging adjacent bits if they are within certain range of each other. in order to avoid that expose properly and overexpose by 1 stop. i do 2 stops since my main field of work is dark dramatic stuff.

@antonstafeyev3606

2 ай бұрын

@@deadendboredom Another reason why Long GOP performs better visually in a static shot is because it compresses based on information from multiple frames. So consider it a NR but the goal is to merge pixels.

Where's part 2

@deadendboredom

Жыл бұрын

kzread.info/dash/bejne/k5mJr9h-htq6pso.html

Are you able to do online paid tutorial on using FX6?

@deadendboredom

Жыл бұрын

Yes, i can do a personal online course on the FX6, just come by on my Discord, or write to me on Instagram.

All-Intra is a much better codec for post-production. The editing workflow is super smooth without the need to transcode in heavy projects unlike Long GOP. The image quality itself is similar, there’s no noticeable difference without zooming on the image, and even then most people can’t tell which is which. The editing workflow alone justify to shoot All-Intra.

@deadendboredom

Жыл бұрын

Thank you for repeating what I said in the video. All-I is better for editing and if you don't need this detail quality, then you have implemented the statement from the video "do your own tests and decide whether this is relevant to you or not" very well. You have decided that it is not relevant to you? Don‘t worry, be happy - but believe me, there are people for whom this information is very relevant. Just to clarify, because I actually didn't put it directly like this in the video: I will continue to use the ALL-I codec for projects that don‘t need the absolute best quality, but for more important productions I will rely 100% on external recordings. I don't want to tell you what to use, I want you to know your tools with their advantages and disadvantages so that you can make the best decision for your work. I'm sorry if I didn't make that clear enough in the video. P.s. Other cameras don’t have this problem with their ALL-I codec and I hope that Sony will improve its codes in the future to reach the level of other manufacturers.

@lukefoxwell

Жыл бұрын

Meh never really had any problems with my editing workflow with long GOP formats. My wallet on the over hand took a beating when I got a PC that would let me have no problems with editing Long GOP!

@vsc604

Жыл бұрын

that may be the case 5 years ago. for $2000 u can buy a base model mac studio, i can do a 4 angle multicam edit in 4k h265 with no stutter, i can even add several grades to my footage with the rendering turned off and it's still smooth as butter

@deadendboredom

Жыл бұрын

@@vsc604 Lucky you, Unfortunately, it doesn't look so good on the Windows side. If I see this correctly, even with the RTX3090 only h.264 420 8bit is accelerated, in h.265 it looks way better. For example, I can't even scrub smoothly through a single XAVC S 4K h.264 422 10bit clip here. Here I would be interested if 4 Long GOP clips work for you at the same time, how much more would it be with ALL-I ;) Currently, with way less than 1000 bucks, I can get the best quality with an external recorder plus super edit performance on all platforms. What i would prefer above all: smooth editing performance for 0 bucks… As I said, I am happy that you can easily and smoothly process Long GOP material, not everyone can. I therefore hope that you also support my desire for Sony to improve their ALL-I codecs so that everyone can work smoothly. 400mbit ALL-I should be feasible, the others can do it too.

Just ran a test myself, yeah there is blocking but nowhere near yours. longGOP look more like what you have recorder in XAVC-I

@deadendboredom

2 ай бұрын

Are you sure that you selected the highest possible bitrate? Perhaps you used a scene with a lot of detail and movement for testing; in that case, XAVC-I is indeed better. Feel free to watch the second video, where I provide many more examples.

@antonstafeyev3606

2 ай бұрын

@@deadendboredom Looking at the video you are shooting with XAVC SI which is a h.264 wrapper. could be an issue, but it looks like something went wrong with the compression and it looks like a blocked noise which never accured. I have never visually seen such an artifact on my fx6 while shooting XAVC-i. And yeah coming from R3D or BRAW world this freaks me out when you sometimes spot a block on extreme grades. However, tracking this footage is meh... while keying is fine, still gona do mattes and stuff so compression will not be noticable there.

All-I is way better quality for me I’ve tested compared to long gop

@deadendboredom

Жыл бұрын

Great that you did your own test to be able to decide what is most suitable for you and your work. But to clarify, because for me semantics are very important here: you did a test and found the "quality" of the ALL-I overall workflow to be better (that's where I'm completely with you), or did you find out that the image quality is better with ALL-I (and if so, is it because you film projects with a lot of movement and Long GOP is therefore not as good)?

Is this a problem with FX9?

@deadendboredom

Жыл бұрын

I have not yet been able to do a test with the FX9, but the specifications say that the bit rate is the same as the FX6. Therefore, my assumption would be that it is also affected - but that is a pure theory. If anyone has the opportunity to take test shots with the FX9, that would be very helpful.

@andrewtaylor5103

Жыл бұрын

It is not a "problem" to start with. If you think it is a problem, and opt to shoot Long GOP without good reason, trust me, you're far more likely to encounter an actual problem.

@deadendboredom

Жыл бұрын

@@andrewtaylor5103 I didn't say that I recommend recording Long GOP for no reason, quite the contrary.

@aleksanderhaugdal2060

Жыл бұрын

Yep

Is this also a problem with fx3?

@deadendboredom

5 ай бұрын

Whether it's a problem or not depends on your requirements; the potentially visible blocks are present in the FX3 just like the FX6 and all other cameras, with correspondingly low bitrates.

@fotografkennethlund

5 ай бұрын

Thanks - oh, so it's not a technical problem but "just" a matter of (too) low datarates ?@@deadendboredom

@deadendboredom

5 ай бұрын

@@fotografkennethlund Yes, I'm sorry if I didn't communicate that well enough; perhaps you should also check out the second part, where I compare more examples in various situations.

@fotografkennethlund

5 ай бұрын

No problem - I really appreciate you taking time to do the video in the first place :) So the take-away is: if we want better image quality, go for PreRes on an external recorder lige Ninja V, right ?@@deadendboredom

@deadendboredom

5 ай бұрын

@@fotografkennethlund I would say there's a two-step takeaway, which depends on whether you only want to record internally in the camera or if external recording is also an option. 1. Technically, external recording in RAW or ProResHQ is the higher-quality recording option. 2. Depending on the recording situation, an exclusively internal camera recording in All-I is not necessarily the highest-quality option. Don't be afraid to record in Long-GOP if it suits the situation. 3. Whether all these findings are relevant for you should be tested out for your recording situations.

sure, but it does not matter for social media content

@deadendboredom

Жыл бұрын

True. I don't see the FX3 or the FX6 primarily as a tool for social media content, but you're right, in that case it wouldn't matter. It doesn't even matter for normal KZread content, I‘d suspect. But wouldn't you agree with me that you should be informed about advantages and disadvantages to then be able to decide for yourself whether you are affected or not? I don't know why, but I somehow expect more in terms of quality from a camera in a „cinema“ line…

@petrub27

Жыл бұрын

@@deadendboredom cinema? and what movies were shot with those camera? for cinema? it's all marketing imo. I use xavsc-i because its smoother on edit

@deadendboredom

Жыл бұрын

@@petrub27 I wouldn’t record internally for the big screen either, because of the problem mentioned. I also feel these labels are more of marketing, in the end every tool used correctly should lead to the desired result, regardless of whether it says „pro“ or „cinema“. If this klickbaity thumbnail bothers you a little, then I will be happy to make changes. To explain: it is important to me that this problem becomes known and that Sony works on its codec - at best so that it improves quality. The over five-year-old GH5 has a better codec.

@johnholidayVFX

Жыл бұрын

@@deadendboredom Big movie from Disney releasing in theaters September, 100% shot with fx3. Sci-Fi

@deadendboredom

Жыл бұрын

@@johnholidayVFX It seem highly unlikely to me for various reasons, but not impossible. Could you please share the name of the movie and the source of the information?

You have totally misconstrued what Sony told you about All I vs Long GOP

@deadendboredom

5 ай бұрын

enlighten me!

This young scientist has not yet discovered that LONG GOP is 8bit 4.2.0. and XAVC I is 10bit 4.2.2. This turns his video into clickbait trash. Nor does he take into account the advantages of shooting with a codec as robust and efficient as XAVC I. Yes, with Prores HQ, you get more "quality", but you pay the price of almost three times heavier files.

@deadendboredom

Жыл бұрын

I already answered the first part under your other comment - if you can't see the difference in quality, then it would indeed be klickbait, but I disagree. By the way, if I remember correctly, I pointed out that this codec also has advantages and that everyone should do a test for their projects and their requirements whether this problem affects them. And because, just like you say, I don't want more than three times as large ProRes files, my wish is for Sony to improve the codec and add, for example, an option with a higher bit rate. I really don't understand why this wish makes you unhappy.

@luissisinno6815

Жыл бұрын

@@deadendboredom Nothing you have said makes me unhappy. Only that your experiment makes no sense, from the moment you draw conclusions about a still image. The same Sony people answered you with some irony: Long GOP might be better for capturing still images, but who captures still images with a video camera? 😂. On the other hand, XAVC I is an industry standard, since Sony cameras are the most widely used in the documentary, television and sport broadcast market.

@deadendboredom

Жыл бұрын

@@luissisinno6815 I simply want a better All I codec and the quality does not improve when the images are in motion. Since an exact repetition of a movement is not so easy, which would be necessary for a good comparison, I did not explicitly test this, but I like to take this criticism to heart. I'm sorry if I went too much into the comparison between IPB and ALL-I in the video, this was mainly used to illustrate the quality difference and thus the lack of quality (For some purposes) of the All I codec. P.s. to be honest, I give a **** on industry standards. Just because everyone says that something is a standard does not automatically mean that it offers optimal quality. Many years ago, DV 720 x 480 was standard and I'm glad that's behind us.

@luissisinno6815

Жыл бұрын

@@deadendboredom The quality of the LONG GOP worsens when capturing moving images. The quality of XAVC I does not improve, but remains standard when capturing moving images, because it captures the whole frame and does so at a higher sample rate and with twice as much colour information.

@luissisinno6815

Жыл бұрын

And my point about the industry standard comes because that standard is achieved with a balance between image quality, efficiency and robustness.