Glidetrack Shooter SD

Just purchased a Glidetrack Shooter SD from Alistair @ www.glidetrack.com. I cannot recommend this product highly enough - it is simply an awesome piece of kit, especially when linked with the Manfrotto 577QR quick release place and Manfrotto 701HDV head.

Here are few videos I have shot with it recently:

http://www.vimeo.com/17981477

http://www.vimeo.com/17979791

http://www.vimeo.com/17953967

1/30s Shutter?

Due to the 180 degree shutter used in film cameras, the optimum shutter speed on a digital SLR should usually be around 1/2n where n = frame rate. For PAL 25fps therefore, the shutter should be 1/50s. If you shoot with a higher shutter than this, you can get very "staccato" looking images, especially during fast motion shots. However, what about shooting at the camera's minimum of 1/30s? The image might become too soft and there might be too much motion blur, however it's something I am going to try soon. The benefit of this is that for low light shooting, you can get away with a lower ISO or smaller aperture.....

Posted a demonstration video showing different shutter speeds at my Vimeo pages: www.vimeo.com/digitalvideoslr

DSLR CAMFRAME

Just bought a DSLR Camframe from B-Hague in Nottingham as shown here (without the camera which was taking the picture!). It provides side and top handles together with two additional hot shoe mounts (which I have used to mount my 126 LED light to one side, and HDMI field monitor / Zoom H4N to the other side).

Canon XLH1s vs. Canon 550D

Decided that in a few weeks time I am going to make a demonstration video comparing the differences / similarities of the Canon 550D SLR and the Canon XLH1s camcorder. I want to compare the DoF available with the same aperture on both. The Canon XLH1s, like many camcorders is around a x7 crop factor so I would have thought that with say an f1.8 aperture, the DoF is still going to be extremely deep. The plan is to film a tape measure from an angle so that the DoF can be measured and compared. The camcorder lens is a 5.5mm to 108mm "L" lens (equivalent to approx. 38mm to 778mm in 35mm format) and has a maximum f number of 1.6. Should be an interesting comparison.

Light tests

Have just taken delivery of 126 LED light. Will post a comparison video of it very soon when I get chance to make one!

Jamie's 30 Minute Meals

Just been watching a behind the scenes video of Jamie Oliver's new Channel 4 series and noticed that they were using a DSLR for some of the shots. Looked like the 70 to 200l f2.8 lens was being used in one of the shots. They also appear to be using RED ONE cameras too.

http://www.channel4.com/food/on-tv/jamie-oliver/jamie-s-30-minute-meals/index.html

LIGHT

Been looking at the Litepanels range of portable LED lights to sit on top of the camera. Their Micro Pro looks excellent but it retails for around £300!

I therefore visited eBay and after a bit of hunting around and review checking on YouTube, I have opted for the "CN-126" 126 LED light panel, which apart from being made from plastic, looks more or less the same as the Litepanels one. The best thing is that it retails for £37.50 and will take rechargeable camcorder batteries.

It also comes with a diffuser filter too and is totally dimmable.

Great Sound Design Website

I always like to show my students sound design / sound for picture "making of" documentaries as I think that they inspire students to start wanting to play with sound and go out and about acquiring sounds which they can later work with.

A student told me about this website yesterday and it simply is fantastic with lots and lots and lots of documentaries regarding sound design for films:

www.soundworkscollection.com

Tokina 11 to 16 f2.8

I think my next lens purchase will be the super wide Tokina 11 to 16 f2.8 lens. It doesn't seem to get a bad review anywhere and it's specially made for cropped sensors. On the 550D, the focal length will be around 18 to 26 which is still nice and wide.

Why a Crop Factor of x1.6 on APS-C sensors?

Just been reading around: (I think this is correct!)

http://en.wikipedia.org/wiki/Crop_factor [accessed 02/10/2010]

http://en.wikipedia.org/wiki/Charge-coupled_device [accessed 02/10/2010]

http://www.december.com/john/photography/cropfactor.html [accessed 02/10/2010]

to try and ascertain a good formula for calculating the crop factor between a full frame sensor and an APS-C sensor (used by 550D / 7D etc.).

The dimensions of a FULL FRAME sensor is 36mm x 24mm with a aspect of 3:2. Dimensions of an APS-C sensor is 22.7mm x 15.1mm, again with a 3:2 aspect.

Square root of (36^2 + 24^2) DIVIDED BY the square root of (22.3^2 + 14.9^2) = 1.6 approx.

This explains why the DoF on camcorders with such a small image sensor (6mm diagonal, 4.8mm x 3.6mm in the case of a 1/3" CCD), are so deep.

The crop factor of this 1/3" CCD sensor would be around x 7. Now, the ratio between each successive f stop is ROOT 2 or 1.414. Comparing a FULL FRAME f stop to an APS-C 1.6 crop shows that the difference in effective f number is around 1.13 fstops (i.e. 1.414 x 1.6). By this what I mean is that f4 on a FULL FRAME camera equates to f2.5 on a 1.6 crop. Now f2.8 would be ONE FSTOP wider, (i.e. 4 / 1.414) so f2.5 is slightly wide than this. Therefore, returning back to the previous example of a 1/3" CCD sensor with a crop factor of x7, how many effective f stops wider would the lens need to be to give the same DoF? Well, 7 / 1.414 = 4.95. Therefore, approx. 5 fstops wider! A 50mm lens with an f8 aperture on a FULL FRAME camera would therefore require 7mm lens with an f1.4 aperture to give approx. same DoF. If we took the same 50mm lens with an aperture of f2.8, we would need an aperture of f0.5???

The 1/3" example above doesn't have the same aspect ratio however, so the calculations will probably differ slightly :)

[Addition]: What about the difference between a x1.6 cropped sensor and 1/3" CCD? Well, as mentioned above, the dimensions of an APS-C sensor is 22.3mm x 14.9mm (diagonal of 26.7mm) and the dimensions of a 1/3" sensor is 4.8mm x 3.6mm with a diagonal of 6mm. Therefore, the crop factor between the APS-C and 1/3" is 26.8 / 6 = 4.55.

It would therefore seem that the effective change in f stops between these two sensors is 4.55 / 1.414 = 3.22 f stops. What this means is that f/8 on an APS-C camera would be the equivalent of a little over f2.8 on the 1/3". Also, to obtain the same field of view on the 1/3" camera, one would need to use a lens 4.55 times smaller (50mm lens, 35mm equivalent = 80mm on APS-C camera and 364mm on 1/3" sensor). So, to give same field of view, I would need to set the lens on the 1/3" camera to 80 / 4.55 = 18mm approx.

A test is going to be carried out to establish whether this is the case, by comparing both the field of view and Depth of Field between a 1/3" video camera and a Canon 550d (APS-C sensor). Watch this space!..... will place a demonstration video on Vimeo.

Here is an interesting article which shows some example shots taken with two different sensor sizes and it sort of agrees with what is being said here:

http://www.janrik.net/DOFpostings/PM1/Depth_of_Field_Versus_Sensor_Size.html [ accessed 14/11/2010].

The crop factor between the two cameras is x3.07. To obtain the same DoF on both sensors, one needed to be set to f/4 and the other f/13. This is very interesting as this represents a little over 3 f stops difference (i.e. f/5.6, f/8, f/11 and a bit more to f/13). The focal length must also have been set to 3.07 times shorter too, although it doesn't stage this in the article. My theory that that the change in f number is equal to the crop factor / divided by 1.414 doesn't therefore ring true here, so it will be interesting to see what my test demonstrates.

[EDIT - 10/01/2010] - JUST FOUND THIS AWESOME FIELD OF VIEW CALCULATOR / DEMONSTRATION TOOL:

http://www.abelcine.com/fov/

DAM - Digital Asset Management

There's not really a right or wrong answer to "looking after" all of the assets which make up a project, but the need for a decent asset management strategy, especially when using double system sound is very important. Students often run into problems when moving from one edit suite / studio to another, and often leave some of their files behind. I have therefore created a basic flow chart which discusses a solution (not THE solution) to help make asset management easier.

This video http://www.vimeo.com/14840608 which I thought was excellent, runs through some really useful methods; my flow chart is attached here too, with a couple of extra stages added if necessary, such as setting up the scratch disc and exporting to ProTools for further sound editing .... (please click on the picture to open it and so that you can read it properly).

The flow chart (and video) makes use of Final Cut's MERGE CLIP feature when using double system sound and this has been discussed and demonstrated on this Blog previously. Equally, PLURALEYES could be used for the automatic sync'ing of clips if a clapperboard of some kind of sync point was not used......

First Electronic Viewfinder for DSLR

Redrock micro have jsut announced the world's first HDMI electronic viewfinder for HDSLR cameras:

http://www.dslrnewsshooter.com/2010/09/26/redrock-micro-launches-the-microevf-for-hddslr/

Monopod

Went out with the camera this morning to a local nature reserve to have a further play with my new 70 to 200L f4 lens. However, this time I also took a monopod and what can I say, these things are great and help take the weight incredibly well. When normally holding the camera with the lens at its maximum focal length, it's so tiresome to hold it steady; not with the monopod!

Great!

Audio drifit issues

Found a very interesting article here discussing a "glitch" in Final Cut Pro which sometimes can cause separate audio to start drifting from the picture:

http://brucesharpe.blogspot.com/2009/06/dslr-dual-system-audio-999-solution.html

I intend to make a demonstration video discussing this shortly.....

ZOOM H4N Output Level Settings

Carried out a little test the other day to compare the output levels of the Zoom H4N when input to the Canon 550D. The audio output is on its own too "hot" for the 550D to handle and as the 550D has no gain controls for its audio input, a PAD cable needs to be employed. Luckily Pink Noise Systems manufacture a custom mini jack to mini jack cable with built in -25dB attenuator to prevent the input on the 550D from overdriving. I wondered however what effect changing the audio output level on the Zoom H4N would have.

With the audio output on 50 (i.e. half way), there was quite a lot of noise present, but compared to a setting of 75 and 100, it sounded far more transparent without any nasty AGC (automatic gain control) kicking on the Camera. I am not interested in keeping the audio any way as I would use separate sound recorded by the Zoom any way, it's just that I wondered how the quality of the audio recorded by the camera would be affected.

Results here: http://www.vimeo.com/15159957

Zoom H4N Limiter Settings?

Just been playing around with the three different LIMITER settings on the ZOOM H4N. The manual doesn't give any information as to what level the limiter actually kicks in at, so I decided to quickly test the three. It appears that on "LIMIT 1" it kicks in at around -9dBFS, "LIMIT 2" at around -6dBFS and "LIMIT 3" at around -3dBFS. I think to try and keep recordings as transparent as possible, I will therefore use LIMIT 3 from now on. Recording at 24 bits, it is not necessary to get as close to 0dBFS as possible, so as long as sufficient headroom is left, the limiter at this level should very rarely kick in; only when the signal is getting very close to 0dBFS....

Sensor Size vs. Depth of Field

Just had a thought - yes, the sensor size does affect the camera's field of view, but does it also have an impact on its depth of field?

Let's take a 35mm camera (such as the Canon 5D MKII) which has a full frame sensor. If we stick on an 80mm lens set with an aperture of f/4, then the diameter of the iris at this setting would be 80mm / 4 = 20mm. However, with a cropped sensor of 1.6, the same effective field of view would be obtained from a 50mm lens (i.e. 80mm / 1.6 = 50mm). Therefore, with this, the diameter of the iris, again at f/4 would only be 12.5mm, i.e. smaller than for the uncropped sensor. So, to obtain the same aperture diameter of 20mm, the f number would need to be 50mm / 20mm = f/2.5 which is more than one f stop larger.

Looking at it another way, if the aperture of 80mm lens on the 35mm camera was set to 12.5mm (i.e. the calculated diameter of the 50mm lens on the cropped sensor), its f number would only need to be 80 / 12.5 = 6.4.

Or putting it another way, the equivalent f number rises or falls by a factor equal to the SQUARE of crop factor:

f/4 on 1.6 cropped sensor = f/2.5 on full frame

f4 on full frame sensor = f6.4 on 1.6 cropped sensor.

This is very interesting as it means that to get the same depth of field that say a 50mm f1.4 lens would give on a full frame sensor, one would need a 31mm lens with an aperture of approx. f0.55 (1.4 / 1.6^2)!

It can therefore be seen that using a full frame camera yields not only a change in the field of view, but also a decrease in the depth of field.

The Road to Coronation Street

Found an interesting article discussing the recent BBC4 Drama "The Road To Cornoration Street" in that it was mostly shot on the Canon 5D MKII Digital SLR:

http://www.4khub.co.za/node/82

Final Cut Issues

Spent an age today trying to compile a demonstration video whereby some footage was 1080p ProRes from the Camera and others were MP4 screen grabs from Quicktime Player X. Firstly, I was unable to get the screen grab footage to load into Final Cut Pro, so I tried transcoding it to ProRes and HDV but neither of these would still load into FCP? Any way, eventually, after using EXPORT TO DV within MPEGStreamClip, I managed to get some footage to load. However, I didn't realise until after several exports of the finished edit, that my sequence preset had been set to PAL DV 720 x 576 due to the format of some of the clips being brought in. This was fine, or so I thought, because I could simply amend the sequence preset from within the settings window. After having changed the sequence preset back to ProRes(LT) 1080p 25fps, the strangest thing happened to all of my footage in that its dimensions drastically reduced, even for the native 1080p footage from the camera? To cut a long story short, the only way to get round this was to re-add each of the pieces of footage back onto the sequence? Very strange??

Vimeo

Have just registered with Vimeo so will be posting higher quality videos there too from now on....

Home page:

www.vimeo.com/digitalvideoslr

Channel:

www.vimeo.com/channels/audiovideotraining

Lilliput HDMI field monitor

Another new toy for rig (well toys as I have also upgraded my Tripod Head to a Manfrotto 501HDV) is the Lilliput 668 HDMI field monitor.

There are many very expensive HDMI capable field monitors out there, but this caught my eye because it was so much cheaper than all of the others. It came supplied with everything required including mini HDMI to HDMI cable, ball levelling hot shoe adaptor, sun screen and battery. I highly recommend this little screen and is great for reviewing footage on and for focussing. Its resolution isn't true HD (800 x 480 pixels) but it more than does the job for me.

See it here:

http://www.carcomputer.co.uk/shop/monitors/hdmi-monitors/lilliput-668gl-70np-h-y-7-hdmi-field-monitor-non-touch-screen

Canon 70 to 200 f4

Just bought a new lens - the Canon 70 to 200 L F4. Absolutely delighted with it and yesterday I carried out a test to establish how the lens performed under different aperture and shutter speed settings. The lens isn't amazingly fast being f4 only, but it is a constant f4 throughout the zoom range and its image is razor sharp.

The test settings used were a constant focal length of 200mm and wide open at f4. As I was shooting in daylight the only way to avoid over exposure was to increase the shutter speed to 1/400s. This is not ideal and I have ordered some cheap ND filters to help bring the useable shutter speed down to 1/50s (for 25p) but these haven't arrived yet. Any way, I then lowered the aperture down to f5.6, f8, f11, f13 and f16, changing the shutter speed to compensate. Interestingly, the "proper" shutter speed of 1/50s could only be obtained with an aperture of f13.

Video here....

Slow Motion

Shot some 720p video using the NTSC format the other day (so that 60fps) can be used. Clip captured and transcoded to ProRes(LT) 1280 x 720 @ 60fps and then slowed down to 42% (i.e. 25 / 60) of its original speed in Final Cut. The "frame blending" option was also unticked.

Example here....

Slow shutter speed, in high light with Shallow DoF - how?!

Recently I went to film a Lancaster Bomber on a Taxi run and I wanted a short DoF, setting my lens to f2.8 or sometimes f4. It was daylight, so light levels were high and due to the wide aperture setting of the lens, the only way to obtain the correct exposure was to increase the shutter speed. Everything looked fine apart from the fact that the propeller blades of the plane appeared to be almost stationary; the high shutter speed of (probably) 1/1000s saw to that!

The problem with shooting at a high shutter speed is that the it sort of takes away the "film" look of the footage, where the shutter speed is always set to TWICE that of the current frame rate being used (as each frame is only exposed to light for half of the time). Therefore, at 25p, a shutter of 1/50s should be utilised. Similarly, when shooting at say 50p, a shutter of 1/100s should be used. So therefore, if you want to use a wide aperture setting for shallow DoF, gain the correct exposure and only use the optimum shutter speed, how you can achieve this; there are no more variables to play with!

The answer (after some research!): some kind of Vari-ND filter attached to the lens which enables the user to control the amount of light actually entering the lens to begin with. Philip Bloom, and others, recommend the Singh-Ray filters (http://www.singh-ray.com/varind.html), but these are very expensive. Therefore, I have found a DIY alternative whereby two polarising filters can be bought and one reversed: http://www.petapixel.com/2010/08/27/how-to-build-a-cheap-and-simple-variable-neutral-density-filter/

Not sure how well it works, but worth a try :)

Rode NTG1 vs. Rode NT1 microphone test

Knocked up a quick demonstration video to compare the tonal difference (if any!) between the Rode NTG1 shotgun mic and the Rode NT1 studio mic. Obviously there are differences in the pick up response between the two mics and I wouldn't want to use the NT1 outdoors, but inside and on axis, they both sound good. There is slightly less "room noise" on the NTG1 but this is down to its hypercardioid pickup. Didn't mention in the video, but both mics were approx. 18" from mouth.

The video itself was shot with the 550d camera and Tamron 28 to 75mm f2.8 lens to give nice shallow DoF. Audio was handled by Zoom H4N and sync'd up with picture using Pluraleyes software (although a hand clap and Final Cut's MERGE SEQUENCE function would have worked just as well).

Exported from Final Cut with a 2.35:1 "widescreen" filter applied. Black bars removed by exporting to MPEG4 (3000kbps) using a resolution of 1920 x 816 (i.e. 2.35:1) and aspect ratio maintained by "crop" option.

RIGS

Really getting into DSLR video now and all its toys. I have been looking into supports and rigs for a while now and whilst the likes of Zacuto are extremely good, they attract a premium. Their "Striker" retails for about £800 alone which is too rich for my blood.

I have therefore been looking at the "Half Inch Rails" website which details more of a DIY approach but who sell really well engineered and light kits. For a complete solution, together with rail mount for an HDMI field monitor etc., and a simple follow focus which can be accessed via your thumb (so no need to move hand away from grip), you're probably looking at about £600, which whilst expensive does allow the option of quickly moving the whole rig onto a tripod. For this, the Manfrotto 577/501 plate combination is recommended. I am looking at the possibility of a 501HDV video head for my tripod, together with a 577/501 plate for the rig so that they can easily be interchanged.

I am also looking at another new lens for my camera and for me, I think it's got to be the Canon 70 to 200 f4 L lens. I cannot find a bad review about this lens any where and, with the cropped sensor of my 550d, would give a zoom range of about 112mm to 320mm and a constant aperature of f4. Ideally, the f2.8 version would be better, but this costs much more and weighs a lot more too! The IS (image stabalised) version is also recommended but again, I'd rather not spend £1000 at least on a lens! Therefore, I think the non-IS f4 lens will be great for me.

The whole rig, when finished, should comprise the Half Inch Rails rig, together with Zoom H4N for audio, some kind of HDMI field monitor and tripod. I already have the LCDVF viewfinder as well.

Just need that lottery win :)

Canon EOS Utility for Training Purposes

Ok, it's been a long time since my last entry, however I plan start blogging again.

I have recently been playing around with the Canon EOS Utility software which came with my EOS550D and it is fantastic for demonstrating lens parameters in realtime, so for lectures and training videos it works a treat. You can use the utility to control all parameters of the lens and camera such as f stop, shutter speed and ISO, together with a realtime preview which in realtime shows what effects changing these have on the image. A cracking tool for demonstrating basic lens theory.

Pluraleyes

What's nice about the Pluraleyes software is that it allows for the audio device to keep rolling when the camera is stopped and started. What I mean by this is that due to the 12 minute approx. recording time of DSLR cameras, these obviously need to be stopped and started. Usually, the audio device would also be stopped and started at around the same point and then the scene would be re-marked again using a Clapperboard etc. The nice thing about Pluraleyes is that the audio file be continuous and it just fits in the video where necessary. Therefore, you can start the audio running, start and stop the camera as much as you like and then throw all of the short video clips at the software, along with the continuous audio file and Pluraleyes aligns all of the clips to the appropriate place along the audio file into a new FCP sequence. Video tutorial to follow soon....

LCDVF

It is my intention soon to get back into blogging and will be posting my findings / tests / videos of the Canon 550d DSLR camera together with other features such as recording double system sound via a Zoom H4N and Final Cut / Pluraleyes etc.

In the meantime, I have just purchased a viewfinder which adheres to the back of the camera via 4 pretty strong magnets and also offers a 2X magnification. I have to say that this device is simply fantastic and provides not only an incredible unobstructed view of the LCD which is essential for focussing (especially outside in sunlight), but also an extra point of contact for increased camera stability.

Here is a picture....



More to come....

Pluraleyes

Very recently discovered some extremely effective software for automatically aligning audio to video when using a double system setup. "Pluraleyes" is a plugin for Final Cut Pro which basically lets you throw any video files at it, together with the accompanying audio files and it will align them automatically and then place each of the newly aligned clips into separate sequences. There is no need for any kind of visual / audio cue (clapperboard for example) and it seems to work extremely well in the basic tests I have done with it.

I will be posting a tutorial video on using this piece of software very soon....

UPDATE: Here are a couple of tutorial videos...

Canon EOS550D

It's been a long time since my last Blog, but I just thought I quickly mention my new toy - the Canon EOS550d DSLR camera. This camera is simply awesome and allows for full HD video at 24fps and 25fps (plus others too). Audio capabilities are not so good, so I have also got hold of a Zoom H4N audio recorder to do double system sound.