Видео

I'm currently working on a new series coming this fall :)

Glad to hear you've found a love in sports photography!! Thank you for the kind words <3

@@TheBiteShot Just look WHO replied!!! My God! It's so kind of you! Ya, actually I had to make that transition. Food photography market in India has got saturated, thanks to the free work delivered to those restaurants by those food vloggers. So, I thought it to be the right time to pivot to a genre that I knew least that i loved but it really makes me feel fulfilled. I love playing soccer, my childhood passion. I thought why don't I mix my present passion (photography) with my childhood one!. And thus.... But wait, didn't I subscribe yet?... Done.

I probably do 4 or 5 ice cream shoots a year

It's listed out in the description box of this video. Enjoy!

Sometimes in the world of video, you can't help but take the ISO fairly high because your shutter speed is locked based on your frame rate and aperture might be limited based on the desired depth of field and maximum aperture of the lens you're working with. So unless you can add a brighter light, ISO is the only option.

If you right click on the file and select properties it should be included there.

FOLLOW UP COMMENT 1 SPEED To understand "speed" we need to know that there is a tiny computer - I/O controller - running firmware in the card, and this has its own piece of extremely fast memory buffer, called "cache". In the old hard drives this had the same architecture. Manufacturers put a "burst" speed on the label of their products and for "write" operations, the simplest way to explain is that a burst must be smaller in size than the cache memory. The I/O controller needs to move your data from the cache to the memory cells for longer and safer storage. These cells are significantly slower and in the old hard drive, the speed at which data could be stored in the magnetic platter(s) - coming from the cache - was called MEDIA SPEED. I will use this term "media" to reference the majority of the card's memory cells, thus excluding cache. Burst speed is interface-controller-cache speed. Media speed is between cache-controller-media and can be sustained until all media have been filled. WEAR Media memory cells are worn every time they are erased, formatted, written. A rewrite should count as one wear event. The price we pay for a card depends on the quality of the media memory cells and their ability to resist wear. The industry has two ways to express cell quality and both refer to how many times a cell can be rewritten. Engineers will use the term P/E cycles (program/erase) and this expresses the actual rewrite number considered a safe upper limit. Most manufacturers are mystical about this and if they publish a number, then it is TBW - Total Bytes Written. Your card may have a capacity of 1TB (1,000 GB) and may have a TBW of 1PB (1,000 TB). This means that memory cells have a life expectancy of 1,000TB/1TB=1,000 P/E cycles - can be rewritten 1,000 times. WEAR & HEALTH MANAGEMENT Proper cards - controllers and firmware in cards - do something called "wear levelling". This is simply a way of spreading data in media memory so it spreads wear evenly. The firmware reports that file A is at the first position in the card, but it may actually physically be in the end of media memory. When you "quick format" a card, then this erases the table of where the files are, as seen by the operating system, but the card is not formatted and the controller knows which blocks of media memory cells have been used how many times. A low level card format resets all media memory cells to 0 and this means that each used block gets another P/E cycle added in the controller's wear levelling administration. How does wear levelling work? Imagine you have a 1 TB (1,0000 GB = 1,000,000 MB) card and your studio shoot typically involves 100 shots of 50 MB. After the shoot, you copy to the computer and backups. Next quick-format the card. After session 1, there's 5,000 MB of blocks used one time. The next shoot, you fire another 100 of 50MB that the controller places after the first 5,000, so now you have 10,000MB of media memory used one time. After 200 shoots, you have used all 1TB media memory once. It may not end here - see over -provisioning. P/E CYCLES LIFETIME EXAMPLE We can compare the media memory cells in our cards to cells in SSD and with SSD it is more common to publish TBW. For example my computer's SSD are the PRO version of one manufacturer's SSD and at the time I bought these, their TBW implied 3,000 P/E cycles while their EVO version's TBW implied 300 P/E cycles. My CFexpress Type B cards have 30,000 P/E cycles and the brand can go up to 100,000 IIRC. OVER-PROVISIONING There are two ways to sell more P/E cycles: (1) use better (more expensive) media memory cells, or (2) over-provision. In the case of (2), there's more media memory in the card than visible on the label. In data-centre-class SSD, these two are combined into over-provisioning of the best cells. The Wear & Health Management example of 200 shoots using a 1TB card for 1 P/E cycle in the case of over-provisioning would mean that the controller has one net label-capacity - see Capacity - with 1st time use. Each shoot, there's 0MB on the card, and imagine the manufacturer over-provisioned 50% under the hood, then this adds another 100 example shoots until the card's media memory is used 1 time. With a very cheap card's memory cells that, say, have 200 P/E cycles and 1 TB is over-provisioned by 50%, this means that such a card could last 200 cycles of 200 shoots.

FOLLOW UP COMMENT 2 VIDEO HOUSE USAGE & WEAR Imagine you are a professional video house and shoot everything at 8K. Now a 660GB card can hold about 3 hours of video. If this card has 50% over-provisioning then a one day shoot that fills the card completely has filled 2/3rd of the under the hood volume. The second day however already starts the second media memory cell wear cycle - P/E cycle. After 3 days, all cells have been used twice. Potentially, depending on the card's firmware, the card is faster when you perform a low level format, potentially it heats less next time, but this adds P/E cycles to blocks. You would probably use more cards to enable you to hand takes over to the post office (office of post processing) and start ingesting takes and sorting them out for the montage and rendering. Say you buy into 256MB card that can hold about 1 hour of 8K. Now you can imagine how fast cheap cards will go through their lifetime expectancy in P/E cycles. BUYING 2ND HAND CARDS If you understand the P/E cycles, then you may be reluctant to buy 2nd hand. I would only do so if the vendor allows me to see the health of the card before I buy. Personally, I would never buy pre-used. Some software exists that can look into the card's wear levelling database and derive a health indication from that. It simply inspects the "secret/hidden-from-you" P/E cycle database that counts the number of P/E cycles for each media memory block. CAPACITY As we have seen above, a card's capacity on the label should be available to the operating system of your computer or firmware of your camera. That does not mean that 100% of that capacity is available to your images or videos. When the card gets formatted, its space is logically divided into space for your data versus space for file names and other things visible to the operating system/application/firmware. After formatting you may end up with between 80% and 90% of what is on the label. This is the net capacity. But there MAY be more under the hood as a way to raise the number of P/E cycles (lifetime relative to wear). DIGITALLY "WEIRD" CARD CAPACITIES Numbers of memory cells are based on powers of 2. where the "power" is an "integer" (whole) number. This gives number series like 16, 32, 64, 128, 256, 512, 1048. In this light, what does 660GB mean? My wild-assed guess is that this once was an over-provisioned card of 512GB label capacity and 512+256=768GB under the hood. As the manufacturer gets zero complaints about this card's lifetime in P/E cycles, they re-release the same product with a new SKU where 50% of the overprovisioning got shifted to the label level. Now the user/operating system/firmware level usable capacity of 768GB under the hood becomes 512+128=640GB with capacity statements one time being based on decimal, other times binary numbers, this is the 660 ballpark. We (may) still have 128GB under the hood as overprovisioning. But the price of the card is adapted to the 660GB capacity. WHAT SHOULD YOU BUY IN P/E CYCLES (TBW) In your computer, where SSD are involved, if you can organise how your data lands in storage, assuming more than one SSD, you can do with low P/E cycles (TBW) in all cases where you store data once and from there on only read them. Your operating system and applications are an example of this. But also your raw images that you store and never change. Something that would cause high-frequency data updates could be your operating system's "page file" or "swap file" that holds a copy of what is in memory and in the way over-provisioning gives more P/E cycles, extends the use of RAM with "virtual memory". You may be physically using 32GB in your 64GB RAM, but the page file may contain 48GB, meaning it has 16GB of data that is not physically in RAM. This can change continually and e.g. in Windows you could remove this wear from your C: drive SSD where Windows normally installs, by assigning the page file completely to a separate SSD or HDD. WHEN I STORE MY CARD OFF-POWER, HOW LONG ARE MY DATA SAFE ON IT? (DATA INTEGRITY and SPONTANEOUS BIT-FLIPS) Ask a team of Ph.D people in electronics, electrical engineering, and quantum physics and they may agree on "7..7,000 hours". One Ytoober-fluencer in a video of a couple years ago, rapped of his memory card strategy: buy the cheapest largest capacity cards, fill with your images and takes, and never erase or rewrite. As he runs a business, buying cards is tax-deducible, he added. This ignores the problem of "SPONTANEOUS BIT-FLIPS" where a 0 becomes 1 or a 1 becomes a 0. That's no big deal when it happens in an image file, but when it happens at the level of the "file system" (see Capacity), it may corrupt that file system and with that the entire card. In Enterprise-Class, Data-Center Class, data storage arrays, the controller can run a frequent process of "data scrubbing" where bit-flips are detected and fixed. That however requires more than one copy of the data and benefits from storing the data in different formats that both allow verification of bit-flips. A very simple approach to such storage is RAID 5 where you have two copies of data in an array of multiple drives. At its base, this is a RAID 0 (striping) array where data are stored across more than one drive (as a way to speed sustained I/O up towards interface/controller/cache speed, away from media speed) and in the case of RAID 5 the data are copied into a second, different, format that gets stored on another member of the RAID set than the primary copy. A RAID 5 controller must do the arithmetic of generating the second copy, and write twice, so part of the RAID 0 gain is lost again. If a drive in a RAID 5 set fails, then the RAID 5 system can rebuild the lost data on a new drive that replaces the failed one. Some memory cards have a data integrity protection system and others don't. How this works and what the level of guarantee is? I don't know, but I prefer to buy cards that promise me they have it.

FOLLOW UP COMMENT 3 WHAT SPEED CARD DO YOU ACTUALLY NEED? Some commenter opines that you don't need a fast card for food photography in a studio. There are always exceptions. For example if you want to shoot in "focus stepping" mode where you generate more Depth of Field by shifting the lens's focus in steps and each time shooting a frame. These are focus stacked in post and in that process, "sharp" pixels are combined into one image. The post software may need to align bits at that first - your lens may have focus breathing where the focal length changes - law of physics/math -when you change focusing distance. You may exclude "architecture" generally shot fro tripod - and again, if you run your sensor in "photosite-shift" mode (generally called pixel shift but a sensor has no pixels) as a way to get more colour space, more colour accuracy in the Bayer paradigm, and more resolution, again post processing aligns and stacks the member shots, then you really want as fast as possible cards to keep the time parallax between the member shots as small as possible. OTHER "QUALITIES" - OPERATING TEMPERATURE RANGE OF CARDS Look for a video by Morten Hilmer shooting musk oxen in Norway in winter with his mirrorless camera. In one of the takes, we see a thermometer indicating about -20C (-4F). Well, most cards are guaranteed down to 0C (32F). The camera he uses in that shoot was also only guaranteed to work down to 0C, but worked excellently. In order to prevent distortion from thermals, Morten keeps his camera and lenses in the cold all the time. This also avoids condensation, e.g. of warmer ir in the tent on the colder camera/lens. I buy cards that are guaranteed to work down to at least -10C (14F). Cards may also differ in upper temperature working range and when a card gets too hot - because you use it continually at high speed for a long time - then it reports that to the camera that either halts or slows down as mitigation of over-temperature with the risk of losing data integrity. OTHER "QUALITIES" - RADIATION, WATER, DUST Some cards report to be able to maintain data integrity when X-rayed, or that their outside is UV-resistant. Frequent security checks that X-ray your luggage and you want this quality. While you will not keep your cards in the sun unprotected, probably, it may accidentally happen in a moment when you swap cards to quickly continue to work. Some cards are better at preventing issues from water or dust or the combination. This is a matter of seals and the material of the outer layer that forms the card's package. One brand specialises in having a metal layer on the outside as extra ruggedness. This may be good for you when you use SD cards that you keep unprotected in your trouser pockets. CFexpress Type B cards have alevel of ruggedness of themselves that removes the added rugged requirement, if any. My cards are stored in a rugged case that has a level of water proof also. It is supposed to float, even with cards in it. BOTTOM LINE You may get what you pay for and a cheap price may be bad for you in terms of Total Cost of Ownership or data loss. Note that some "cheap" countries of origin may have their own sense of humour. One brand of flash devices in Chinese is called "God Ox" - mistranslation of "Holy Cow"- not Go Dogs. Well, the cheap Li-Ion battery called 3,900 mAh may just be a joke about some Chinese word "ma(h)". A 4TB card for a few bucks may give you 4MB. The problem, even between trusted reputable brands, is that different cards score differently in each of these qualities and manufacturers/vendors/brands are mystical about them. To one vendor of very interesting cards - speed and capacity at a low price - I asked about these other qualities. They answered my P/E cycles (TBW) question with "enough". Commercially, they had given me the middle finger and I will never ever buy that brand. The cards I now standardised to, are from a brand that politely answered all my questions in this regard, at a deeper level actually than my questions, in terms I could understand.