Luminous Landscape Forum

Equipment & Techniques => Motion & Video => Topic started by: BernardLanguillier on February 26, 2012, 05:17:31 pm

Title: Hyperdeck Shuttle review
Post by: BernardLanguillier on February 26, 2012, 05:17:31 pm
Thanks Michael for the review of the Hyperdeck Shuttle.

Two questions coming from a person with little video experience if you don't mind.

My understanding is that an HDMI signal can only be considered as clean if 2 conditions are met:
- zero signal compression,
- no text overlay (this being less relevant in the context of your article since it focuses on image quality)

1. Based on your experience with video, how does the compression of signal show up in files? I mean what are the areas where a non compressed signal is expected to make a difference? Guessing from my still experience:
- lack of edge artifacts?
- better ability to post-process the files (slow motion generation,...)?
- better ability to display on large screen?
- ...

2. Your article starts with a mention that quite a few video DSLRs are able to output a clean HDMI signal. I am not sure what you mean by "video DSLR". I guess that it probably includes both motion cameras with interchangeable lenses and still cameras able to shoot video? As far as the latter goes, it seems that as of now the only two cameras having been announced that are able to output a real clean HDMI signal (non compressed and no overlay) are the Nikon D4 and D800.

Once they become available, do you expect to have the opportunity to check whether the Shuttledeck makes a difference in terms of image quality with either of these 2?

Thanks.

Cheers,
Bernard
Title: Re: Hyperdeck Shuttle review
Post by: michael on February 26, 2012, 06:09:58 pm
Bernard,

Compression can display itself several ways. So-called macro-blocking is one, where the image is changing so fast due to subject or camera motion that small squares (artifacts) are generated. There is mosquito noise, which is a kind of shimmer around edges, and quite a few others as well.

I really have no firm info on which video-capable DSLRs and CLCs output uncompressed video through their HDMI ports. A bit of web searching should uncover which ones do and which don't.

Nikon has promised an D800 for testing (not sure if an 800E will also be available at this point). Regrettably they tend to be somewhat slow in following through on test samples though. If not I'm sure that I can borrow one from a dealer, though this will have to wait till I'm back in Toronto in mid-April.

Michael
Title: Re: Hyperdeck Shuttle review
Post by: fredjeang on February 27, 2012, 04:00:51 am
Hi,

I must say that I've liked the tone of this article. Because I think that it's absolutly necessary for the reviewers-testers-field testers (or call it as you want) to have from now a cautious and contrasted message when it comes to those dslr-evil cameras for motion. There is an enormous amount of marketing everywhere and unlike stills, it's not always as easy to find quickly (and I recognise the difficulty also for the reviewers) the issues that will show-up.

The tone of this article inciting to find the more possible infos on-line, read as many articles as we can and making references to external sources is to me the right message to transmit. Specially with motion and specially with dslr motion.

The problem is that we are in a singular and transitory situation. Until Raw video is not becomming the norm, like in still imagery, and by norm I mean reliable and accessible for all the market and not just like now the high-end expensive professional, and until the cards capables aren't cheaper (that's the big prob of the SSD), the investment in Dslrs-evils for motion has to be carefully meditated because there are hassles (a part from the design itself) and useless expensive purchases that won't have any significant impact on the results.

No, we're not yet in the situation that we can purchase a Nikon D4 for 3000 bucks, plug it an external recorder with those super-fragile mini hdmi for another 3000 (with "enough" ssd) and magically for 6000 we have an Alexa-on-the-cheap. That's not what is going to happen and people should really thing about 10 times if they're going in serious into motion but on a budget.

I also come to the same conclusion as Michael and many others sources: putting an external recorder with lossless adquisition codec does not improve, for those consummer cameras, the image quality in mormal circunstances and slightly improve it in delicate imagery (lack of light for ex and how the shadows are rendered).
The gain of such recording on dslrs is so marginal and the hassles of the cost and file-size is such (bigger than R3D !!) that it's completly meaningless IMO.

Also, DNxHD is great if you are on Avid, but as the majority of small-middle prod houses and independant filmakers are on FCP or Premiere...

In other words: the image quality that deliver dslrs-evils out of factory is more or less what you should expect to have with little room for significant improvements. Hack or no hack, external recording or not external recording. If you want-need more: I'd stop investing money in those gear and save it for an F3 or a Red etc...




A word on the keying: I understand that Michael wanted to show what can be expected in FCP10 as a based keying. It's ok but obviously far from perfect.
I think personally that it would be, in this context, useless that we jump (like we unfortunatly see too many times in the forums) on this keying and wouldn't bring anything into the thread.
I take it as it, it shows a quick sample of what's to be expected right-of-the-box from an editor. Nothing to bark against in the context of this article and the software used.

Keying is an art and I've seen experienced Nuke operators that are always using 2 or 3 different keying techniques to acheive perfect results and it takes them time...
Again, thinking that with a FCP10 editor style at 250 bucks you'll have perfect keying pressing just a button, is fooling yourself, Apples or oranges. (and I know that somebody would jump on that to show me I'm wrong because they managed to key perfectly at first with it).
It's pretty much straighforward for average results but not reliable enough for high-end assignements. Sometimes the key might work perfectly, luck; but in most of the cases it will require
lots of experience and elaborate keying engines. I see here again another reason to stress to not fool ourselves and keep feet on earth.
Title: Re: Hyperdeck Shuttle review
Post by: BernardLanguillier on February 27, 2012, 05:37:27 am
No, we're not yet in the situation that we can purchase a Nikon D4 for 3000 bucks, plug it an external recorder with those super-fragile mini hdmi for another 3000 (with "enough" ssd) and magically for 6000 we have an Alexa-on-the-cheap. That's not what is going to happen and people should really thing about 10 times if they're going in serious into motion but on a budget.

You may be right, but I guess we will be able to say that for sure after the D4/D800 are released and tested?  ;D

Right now your statement appears to be based on a generic feeling that all DSLRs are identical. We know from our still experiences that this is very very far from being the truth so I am not sure why we assume that this is going to be the case for video?

In fact, I have been unable to find any other DSLR providing a non compressed output through HDMI, so I am not even sure what this feeling is based on.

Cheers,
Bernard
Title: Re: Hyperdeck Shuttle review
Post by: fredjeang on February 27, 2012, 05:55:43 am
Bernard, I don't see it.

You generally are realistic. We don't even need the release of the Nikon and tons of testings, it's a 8 bit camera, no access to DNxHD 444 nor Prores 444, no access to curves import, no raw, no high frame-rate etc etc... it's like someone would sell you a still camera that would only featured jpegs at 7 quality. Very little room in post. Very little options to failure. There is no miracle, even for Nikon. You'll see the banding festival and other compromises.

There is a big difference in the still world (a Raw world) and the new DSLRs video "toys" (toy is not bad, a toy can be of course a tool). Check how an Alexa is built inside and it tells it all.

I know how much you love Nikon, but those dslrs cameras in motion  are good for reportage, war news, for the indy wanabee, for some low-end low-budget Hollywwod Bollywood series, good also as a HD bloc note, third camera, or bring it with you and film when you climb mountains etc...but it's not going to replace a F3, a Red, an Alexa... add any motion models for other brands like Thomson etc...  

This camera will have probably a clean nice look, with all the list of hassles we already seen in those type of gear. No spec is saying that it will be revolutionary nor different, as often with Nikon, a good product within the choice made. I accept much more the drslr-evils limitations from a V1 point because it's really tiny. IMHO.

But, ok, fair enough, let's wait and see the testings. I have zero faith it will be different than what we already have in this segment.
Title: Re: Hyperdeck Shuttle review
Post by: BernardLanguillier on February 27, 2012, 06:20:07 am
I know how much you love Nikon, but those dslrs cameras in motion  are good for reportage, war news, for the indy wanabee, for some low-end low-budget Hollywwod Bollywood series, good also as a HD bloc note, third camera, or bring it with you and film when you climb mountains etc...but it's not going to replace a F3, a Red, an Alexa... add any motion models for other brands like Thomson etc...  

All I am saying is that, specwise, it seems that the D4/D800 may be the only 2 DSLRs for which clean HDMI might be making a difference.

I have no doubt that even if there is difference thanks to this clean output, there will still be an important gap compared to dedicated cameras.

Frankly I am totally new to video and have no particular stakes in the business. It wouldn't hurt me the least bit if Nikon were worse.  ;D

Cheers,
Bernard
Title: Re: Hyperdeck Shuttle review
Post by: fredjeang on February 27, 2012, 06:37:51 am
There are many people currently doing networking and gaining worldwide reputations on-line with what they often call themselves "the new generation of filmaking", something like that. As opposed to the orthodox crews. No need to name them, they are already famous and we know who they are.
The Dslrs-for-motion and the tons of accessories and compromises needed are their main targets because they don't target the high-end professionals, they can't. They sell us the "motion-on-the-cheap" buzz.

Be very carefull with all those "informations". All that is very fashionable. Manufacturers have seen the buzz too, they also use marketing expressions like "multimedia". This is all fake, to some extend.
This is big marketing hoax.


Cheers.
Title: Re: Hyperdeck Shuttle review
Post by: Morgan_Moore on February 27, 2012, 06:47:41 am
Maybe I read the article wrong but it seems to imply a tenfold factor in disc spaced used - surely with the Fs100 25mbs and Pro-res 100mbs thats a factor of four, that considering on the computer you transcode anyway it seems that total HD space would actually be less (of your w/flow is transcode everything).. . in exchange for requiring more memory in the field

It seems that gaining bit depth is a major consideration to using a recorder meaning the file (tonal skies etc) will fall apart less with a recorded master under grading (-maybe sony F3)

If the camera doesn't pump out 10bit then this advantage is lost ?

One recorder does seem to have an advantage in the (D4 D800) workflow and maybe Fs100 etc workflow is the Sound Devices Pix 240 in that it supports Timecode and Presents XLR sound inputs with excellent pre amplification

Timecode is still not clear to me - it seems that there is 'timecode' and 'jammable timecode' the second is actually a solid way of synching multiple recordings, vision or audio, whereas the first may be just pretty looking numbers on the screen (my guess with the FS100) which might look a bit 'pro' to the client

An aside . MR did not mention the ability to edit to deadline using an edit ready codec from a recorder?

Maybe someone can clarify these points

SMM
















Title: Re: Hyperdeck Shuttle review
Post by: Morgan_Moore on February 27, 2012, 06:50:40 am
There are many people currently doing networking and gaining worldwide reputations on-line with what they often call themselves "the new generation of filmaking", something like that. As opposed to the orthodox crews. No need to name them, they are already famous and we know who they are.
The Dslrs-for-motion and the tons of accessories and compromises needed are their main targets because they don't target the high-end professionals, they can't. They sell us the "motion-on-the-cheap" buzz.

Be very carefull with all those "informations". All that is very fashionable. Manufacturers have seen the buzz too, they also use marketing expressions like "multimedia". This is all fake, to some extend.
This is big marketing hoax.


Cheers.

I think this is time for a new topic!

IMO there is both truth and fake to the claims

Truth: You maybe aware of Danfung Dennis nearly (but not) getting an Oscar last night for his film shot on a 5d

Fake: Often use of a dedicated camera e.g. FS100 or F3 is cheaper than a 'tricked out' DSLR

S
Title: Re: Hyperdeck Shuttle review
Post by: fredjeang on February 27, 2012, 06:58:53 am
Maybe I read the article wrong but it seems to imply a tenfold factor in disc spaced used - surely with the Fs100 25mbs and Pro-res 100mbs thats a factor of four, that considering on the computer you transcode anyway it seems that total HD space would actually be less (of your w/flow is transcode everything).. . in exchange for requiring more memory in the field

It seems that gaining bit depth is a major consideration to using a recorder meaning the file (tonal skies etc) will fall apart less with a recorded master under grading (-maybe sony F3)

If the camera doesn't pump out 10bit then this advantage is lost ?

One recorder does seem to have an advantage in the (D4 D800) workflow and maybe Fs100 etc workflow is the Sound Devices Pix 240 in that it supports Timecode and Presents XLR sound inputs with excellent pre amplification

Timecode is still not clear to me - it seems that there is 'timecode' and 'jammable timecode' the second is actually a solid way of synching multiple recordings, vision or audio, whereas the first may be just pretty looking numbers on the screen (my guess with the FS100) which might look a bit 'pro' to the client

An aside . MR did not mention the ability to edit to deadline using an edit ready codec from a recorder?

Maybe someone can clarify these points

SMM


That's where the dilema is. The only thing that you gain is that it bay-passes the compression artefacts of the AVCHD to a less destructive (but still destructive) codec like DNxHD or PRORES 422.
But in how many bits...
The file size is huge for a gain in quality that isn't.

 
Title: Re: Hyperdeck Shuttle review
Post by: Morgan_Moore on February 27, 2012, 07:01:30 am
It would seem that way

I think a time may come when there is say a D5 that can do 11mp NEFs into a recorder - but not yet

S
Title: Re: Hyperdeck Shuttle review
Post by: fredjeang on February 27, 2012, 07:02:30 am
I think this is time for a new topic!

IMO there is both truth and fake to the claims

Truth: You maybe aware of Danfung Dennis nearly (but not) getting an Oscar last night for his film shot on a 5d

Fake: Often use of a dedicated camera e.g. FS100 or F3 is cheaper than a 'tricked out' DSLR

S
Totally agree.
Dslrs gear doesn't mean that somebody with talent, creativity etc...can't do great stuff with those and win an oscar.
But we should be clearly aware of what does it means in terms of hassles, false economies, and limitations and their impact compared to dedicated motion cameras into the business models of each pro.
Title: Re: Hyperdeck Shuttle review
Post by: Tim Jones on February 27, 2012, 10:56:12 am
 I went to a movie this weekend because it was shot on a 5D at a major cinema chain,.  After about five minutes i totally forgot about the camera it was shot with. It was a good movie.
It being shot on a dslr made it different, but in a good way.  You could tell it gave them a lot freedom to try new things .  It was great.
I know it took two years in post to make it look like this though .  So, there are trade offs of course.
Thanks,
Tim
www.tjphoto.net
Title: Re: Hyperdeck Shuttle review
Post by: bcooter on February 27, 2012, 01:06:37 pm
Totally agree.
Dslrs gear doesn't mean that somebody with talent, creativity etc...can't do great stuff with those and win an oscar.
But we should be clearly aware of what does it means in terms of hassles, false economies, and limitations and their impact compared to dedicated motion cameras into the business models of each pro.


My suggestion to anyone making the addition of motion imagery to their still repertoire is use what works for your business, not the art.

The art is easy.  I mean if your going to build a reel, or test, or shoot some psa's, editorial whatever to get your feet wet, nobody will care if it's shot with a RED, an Arri or a 5d2, unless you have some major issue like moire or bad color grading.

The business side is more complicated.  Some client's expect a better camera than a dslr on set, some don't care, but I wouldn't suggest overspending, at least not in today's world.

Having video experience has been good for our business, but it's a double edged sword, takes investment and risk and a huge volume of time.

You also have to be aware that the world of motion imagery has been tightly squeezed.  What use to be a low budget movie at 2 million is now half a million to 800 thousand, sometimes 1/10th of that.

So those directors have moved more into commercials, even very accomplished directors with huge CVs.   In turn that moves the medium range of directors more towards web and display video, which puts more pressure on photographers trying to get a handle on how to profit from motion imagery.

The one thing still photographers offer that the motion guys usually don't is the ability to work with smaller crews, usually with faster on set turnaround.  It's not unique for a still photographer to be part director, dp, camera operator and even part producer, where in the film world, it's rare you'll see a director holding a camera for any period of time.

Regardless, IMO, the best camera for building your reel is a 5d2.  It's fast, it's easy, goes to mega iso, and except for the sound issues and sometimes moire it's a very workable camera and file and lenses are everywhere.

For heavy lifting the air becomes more rarified.

Personally, I've used about everything but the Arri.  Nikons, Canons of all kind,  the sony fs100, even the little Panasonic.  I still think the 5d2 is probably the best due to it's weight, iso and availability and though you need some time of sound recorder, once again it all depends on what you shoot.

For real world work, I love the RED One's, today even better than our Scarlet, because even though the R1 is heavier it's a lot more simple camera to work than a Scarlet.  I assume the Epic is about the same though haven't used one.

Once I use the Scarlet more I might change my mind, but (knock on wood) our RED One's have been very dependable.

In summary, digital motion really does mirror digital still photography.   You can build your body of work with smaller, easier to use cameras and when the time comes, step up to something more expensive in either purchase or rentals.  Just don't overspend, not if your in business.

IMO

BC

Title: Re: Hyperdeck Shuttle review
Post by: fredjeang on February 27, 2012, 03:52:55 pm
In summary, digital motion really does mirror digital still photography.   You can build your body of work with smaller, easier to use cameras and when the time comes, step up to something more expensive in either purchase or rentals.  Just don't overspend, not if your in business.

Yes. It's a sort of delicate equation between not overspending and at the same time avoiding false economies. Not always easy.

I agree that the 5D2 is still probably the most acheived dslr-style camera today bang-for-the-buck, despite its known issues. The very capable GH2 is delicate in use, has some factory problems never adressed and won't be as good in higher isos; now both the Canon and the Panasonic are hackable.

To me, the micro 4/3 for video is a good standart however, probably "better" than the full frame 35mm from CaNikon, but it's a worse standart for still, thinking mixed use. I think that mirrorless micro 4/3 has a great future in motion, it allows to mount much more cinema lenses without body modification nor vigneting, some of wich are high quality for a reasonable cost on the cine 2nd hand market. Also the D.O.F is almost identical to what we are used in our cinema theaters. But they need a more robust product for their next generations and being less protective towards their pro line. The commitments from some major lens manufacturers with this standart shows that a serious part of the industry beleives in it. But for the moment I'd recommend a 5D2 over a GH2 for the reasons James mentionned above.