Hyperdeck Shuttle 2 now does ProRes.

What is the battery life like? I use a Ninja with my FS100 and though more expensive it give ma a second monitor for the client and the battery life is excellent along with being able to use Sony batteries. I do like the SDI feature of the BMD box..
 
Internal battery lasts an hour to 90 min if I remember correctly, I use V-Mounts to power everything so I can't remember exactly. I'd maybe just velcro on these.
 
Last edited:
BrianMurphy said:
What is the battery life like? I use a Ninja with my FS100 and though more expensive it give ma a second monitor for the client and the battery life is excellent along with being able to use Sony batteries. I do like the SDI feature of the BMD box..
Click to expand...

Awful. Although i remember my shuttle v1 being pretty good. My suspicion is that DNxHD and ProRes are draining the battery much faster than uncompressed.
Maybe an hour?

cheers
paul
 
Thanks for the answer. I will stick with my Ninja for the time being it works well .Just a pain to start and stop, though I am working on a fix for that right now.
 
I recall reading in a number of earlier posts that external recorders do not add anything significant to the AVCHD image from the FS-100.

For those working with any of these recording units, what do you find are the advantages of recording on them, beyond saving the producer the time needed to transcode to ProRes or whatever.
 
joe/dp said:
I recall reading in a number of earlier posts that external recorders do not add anything significant to the AVCHD image from the FS-100.

For those working with any of these recording units, what do you find are the advantages of recording on them, beyond saving the producer the time needed to transcode to ProRes or whatever.
Click to expand...

I read similar posts, and frankly most of those people have no idea what they are talking about. Most of them were based on tests where the camera wasn't moving and there wasn't much noise. The primary benefit of external recording to DNxHD/Prores is in scenes with high noise or high detail or uncompressible content. I have hardly used my Shuttle 2 with my FS700, but in a couple quick tests I did to test my SSDs it was immediately clear that at higher ISOs and higher noise levels, ~200Mbps Prores was performing an order of magnitude better than 24Mbps AVCHD.

AVCHD is amazingly good for it's bitrate as long as scene details do not change quickly, but it depends heavily on temporal compression and it cannot handle quick detail changes or noise as those are very difficult to compress within a frame and extremely difficult to compress between frames.

Anyway, for the most part, even for people capturing externally will mostly still be using AVCHD for overcranked framerates. Then again, to some extent, higher frame rate content may be slightly easier to compress given that scene motion or detail changes will be slowed down.
 
Thanks, Noah.

If anyone has screen shots that demonstrate the difference between straight AVCHD shots straight out of the camera, and improved images from those recorded with any external recorder, I would love to see them.

Joe Friedman
DP
www.josephfriedmancom
 
It still won't do the pull down for 24P. I would not be interested in it until it does just one more thing to have to do in post.
 
I just pulled out my camera and Hyperdeck and did this quick comparison, hope it helps illustrate the AVCHD vs Prores question.

http://www.noahyv.com/temp/FS700HYPERDECK2COMPARISONS.jpg

Comparison #1: Broad daylight, static camera, normal settings with moderately detailed scene
As you can see, especially looking in the shadow areas, even in a high key scene with little noise, Prores holds detail and some noise structure a bit better in shadows. AVCHD is a bit muddier. AVCHD does quite well in this situation since its weaknesses are noise and movement, but even here the differences are visible. Notice even here, the edge of the red bricks have better separation in Prores than AVCHD.

Comparison #2: High ISO
Similar results but the loss of detail in the AVCHD is greater due to the increased noise level. Higher noise level forces more detail smoothing for AVCHD, which reduces contrast, especially pixel contrast. Prores holds pixel contrast better and noise structure.

Comparison #3: Extreme low light, Higher ISO
AVCHD does not do well at differentiating between noise and real detail, both are lost. This is a bit extreme as it is high ISO and underexposed, but it shows you where AVCHD really breaks apart. AVCHD acts as a heavy denoiser, which some people may like to some degree, but there is a significant chance of losing real detail and recording a lot of muddy shadows.

Comparison #4: Broad daylight, moving camera, increased shutter
Here is another place where AVCHD falls apart a bit. It is so efficient because it uses similar information in previous frames to define the current frame, but when the frame changes more quickly, the codec displays errors and loss of contrast. There is a lot of mud and odd looking edges where the codec may be trying to keep detail that seems similar to previous frames.

I noticed in the last test that parts of the frame were quite a bit worse, as there the codec would fall apart in even higher motion, but I chose that frame and crop because it was similar to the previous tests. I was a bit surprised how much the codec had trouble with high motion. This was surprising since I figured that frames in motion would be easy to compress since they were quite blurred.
 
Last edited:
Are the differences as pronounced when using the 100mbps flavor of ProRes? I've considered using an external recorder for select projects, but I just can't manage the mountain of data that would be generated at 220mbps, so I've been sticking with AVCHD 2.0 at 28mbps/60fps so far.
 
joker454 said:
Are the differences as pronounced when using the 100mbps flavor of ProRes? I've considered using an external recorder for select projects, but I just can't manage the mountain of data that would be generated at 220mbps, so I've been sticking with AVCHD 2.0 at 28mbps/60fps so far.
Click to expand...

I would highly recommend against using the 1080p60 28Mbps option since it is extremely compressed, giving only 0.46Mb per frame. 1080p24 at 24Mbps gives you 1Mb per frame and is a compression ratio of 24:1, so it is more than 2x less compressed than the 28Mbps flavor which is a compression ratio of 53:1.

I've heard stories of the 28Mbps AVCHD really falling apart.
 
nyvz said:
I would highly recommend against using the 1080p60 28Mbps option since it is extremely compressed, giving only 0.46Mb per frame. 1080p24 at 24Mbps gives you 1Mb per frame and is a compression ratio of 24:1, so it is more than 2x less compressed than the 28Mbps flavor which is a compression ratio of 53:1.

I've heard stories of the 28Mbps AVCHD really falling apart.
Click to expand...

It's not too bad, it really depends on your content. For two of my websites the videos are all filmed via tripod and I post them to the website in 60fps. For the content matter of these two particular websites they just look much better in 60fps, 24p or 30p does not look as good to me, so 60fps recording is mandatory here. There aren't sweeping background movements in these videos either since they are all tripod filmed so the stock AVCHD codec compression works ok for my needs here.

What I do notice though I believe results from the 4:2:0 recording of the AVCHD codec on the camera. When certain colors are next to other colors I find things just breakdown. So if I'm doing a tripod shot of a lady and there is a stationary red rose in the scene against a different color, the edges of that rose will appear blocky. I don't think that's an AVCHD compression artifact as the rose isn't moving at all, but from what I've read it's the 4:2:0 chroma sampling limits coming into play. That's why I started to look into external recorders that can record in 4:2:2, which led me to Prores and DNxHD. But I can't muster the mountain of data they create. On a typical 4 days of shooting I record about ~260GB of data on the stock AVCHD 28mbps codec. That means using the 220mbps flavor or ProRes would leave me with ~2TB of video data. I keep three magnetic backups of all shoots and one optical backup (burned to bluray) all stored at two different banks, so using ProRes 220mbps 4 days of shoots would leave me with ~6TB of magnetic footage and 2TB of optical footage to manage. It's just too much to handle and manage. Hence why I'm curious about the lesser quality flavors of ProRes and DNxHD since it's the 4:2:2 I think I need more than the 220mbps. But I wonder if the 100mbps of ProRes is any good, since it seems like anytime someone uses an external recorded with ProRes they just use the best quality version of the codec. I haven't heard any info on their "LT" version at 100mbps.
 
You must log in or register to reply here.
Back
Top